Good morning! This is your daily ☕️ Techpresso.
In today's Techpresso:
💥 Trump's latest hire is a hit to Big Tech
🤔 Microsoft accused of monopoly tactics
🔥 Nvidia's AI chips face overheating concerns
👀 Wall Street giants prepare for biggest tech infrastructure boom since the internet
🎁 + 6 other news you might like
🔮 + 6 handpicked research papers and tools
💥 Trump's latest hire is a hit to Big TechLINK
President-elect Donald Trump has appointed Brendan Carr as the Federal Communications Commission chairman, a role that does not require Senate approval because Carr has been a member since 2017.
Brendan Carr, who replaces Jessica Rosenworcel, aims to impose new restrictions on social media to favor conservative views and reduce the Section 230 legal shield protecting user-generated content.
Despite ambitions to regulate big tech companies like Google and Meta, the FCC under Carr would need new legislation to expand its powers, as these firms are not classified as communications services.
🤔 Microsoft accused of monopoly tacticsLINK
Microsoft is under investigation in the US for allegedly providing free services to federal agencies to prevent competition and maintain government contracts.
Reports revealed Microsoft pledged $150 million in 2021 to offer cybersecurity services, enabling government access to its Microsoft 365 Government G5 suite.
Microsoft's strategy allegedly made government agencies dependent on its tools and increased Azure adoption, challenging Amazon Web Services' market dominance.
🔥 Nvidia's AI chips face overheating concernsLINK
NVIDIA's new Blackwell chips are facing overheating issues when tightly packed in server racks, leading to concerns about possible delays for this highly anticipated AI hardware.
The company has requested several design changes from suppliers to address these overheating problems, which has added uncertainty to the release schedule.
Though a spokesperson minimized the issue, the need for late-stage modifications suggests possible impacts on upcoming shipments and raises questions among major customers like Meta, Google, and Microsoft.
👀 Wall Street giants prepare for biggest tech infrastructure boom since the internetLINK
Major financial institutions and investment firms are preparing for a significant boom in AI infrastructure, expecting to invest at least $1 trillion in data centers, energy systems, and network setups.
OpenAI's Sam Altman suggests building larger facilities, with some US states planning AI data centers costing over $100 billion each, while BlackRock and Microsoft plan infrastructure investments up to $120 billion.
While financial giants like JPMorgan and Deutsche Bank are rapidly pursuing data center deals, experts caution that much capital might be wasted if AI development does not meet high expectations.
Other news you might like
Google, Amazon and Apple are racing to get AI on their most popular devices.LINK
TikTok parent ByteDance’s valuation hits $300 billion, sources say.LINK
Sam Altman told Elon Musk he worried Google Deepmind was trying to 'kill' OpenAI by poaching its talent.LINK
SpaceX's European Rival Raises $160M For Reusable Space Capsule.LINK
1,900 times Earth’s gravity: China activates world’s most advanced hypergravity facility.LINK
Google’s AI chatbot tells student needing help with homework to ‘please die’.LINK
Latest research and tools
Nova JavaScript Engine: a JavaScript and WebAssembly engine written in Rust, designed for Rust projects to embed and run JavaScript, featuring a data-oriented design inspired by other engines and focused on efficient execution.LINK
It's time to replace TCP in the datacenter (2023): the paper argues for updating the TCP protocol to enhance performance and efficiency in datacenter networks.LINK
C-web-modules: enables real-time uploading and compilation of C code on a server, designed for efficient web development using a modular approach that supports hot reloading and integrates with external libraries like SQLite3 and OpenSSL.LINK
Documind: an open-source AI tool that processes documents into structured data, utilizing OpenAI's API for information extraction and allowing customizable schemas for various formats, suitable for deployment on both local and cloud environments.LINK
Nokolexbor: a drop-in replacement for Nokogiri that offers significantly faster HTML parsing and CSS selector performance, aiming for API compatibility with Nokogiri and providing pre-compiled gems for common platforms.LINK
LLaVA-O1: the paper introduces a method that allows vision language models to reason in a step-by-step manner.LINK
Want to get the latest news differently? Find us on:
See you tomorrow for a new dose of ☕️ Techpresso!
-min.png){kind=small}
.png){kind=small}
